PORT:%20A%20Price-Oriented%20Reliable%20Transport%20Protocol%20for%20Wireless%20Sensor%20Networks

1
PORT A Price-Oriented Reliable Transport
Protocol for Wireless Sensor Networks
ISSRE 2005 Presentation

• Yangfan Zhou, Michael. R. Lyu,
• Jiangchuan Liu and Hui Wang
• The Chinese University of Hong Kong
• November 10, 2005

Simon Fraser University
2
Presentation Outlines

• 1. Introduction
• 2. Design Considerations
• 3. Protocol Implementation
• 4. Simulation Results
• 5. Conclusion

3
Presentation Outlines

• 1. Introduction
• 2. Design Considerations
• 3. Protocol Implementation
• 4. Simulation Results
• 5. Conclusion

4
Wireless Sensor Networks (WSN)

• Sensors nodes measure physical phenomena.
• Target tracking
• Intruder monitoring
• Environment data measurement
• Moisture, temperature
• Engineering or scientific data measurement
• Habitat surveillance
• Sensor nodes form an ad-hoc multi-hop wireless
  network to convey data to a sink.

5
Introduction
Sink
6
Introduction

• Reliable sensor-to-sink data transport for WSN
• It is important as it ensures the mission of the
  networks
• Objective
• To ensure that the sink can receive desired
  information
• The work presented here is to address this
  problem.

7
Introduction

• WSN Challenges
• WSN suffers from energy constraints
• WSN conditions
• Unreliable wireless link
• High and dynamic packet loss rate
• Network Dynamics
• Node failures
• Link failures
• Dynamic traffic load

8
What Is Addressed

• What should a reliable sensor-to-sink data
  transport protocol do?
• Ensure that the sink can collect enough
  information
• Minimize energy consumption of data transport
• How should it be designed to achieve the goals?
• With cooperation of the application layer
• Adjust the reporting rates of sources
• Adapting to wireless communication conditions

9
Presentation Outlines

• 1. Introduction
• 2. Design Considerations
• 3. Protocol Implementation
• 4. Simulation Results
• 5. Conclusion

10
Reliable Sensor-to-Sink Data Transport

• Ensure that the sink can obtain enough fidelity
  of the knowledge on the phenomena of interest
• 100 packet delivery is not necessary.
• The key is that the desired information can be
  obtained.
• Only the application that utilizes the packets
  knows whether the data transport is reliable or
  not.

11
Observations

• Different sources have different contributions to
  improve the sinks knowledge on the phenomena of
  interest (known by the application)
• Different energy is required for communications
  between different sources to the sink (known by
  the transport protocol)

Sink
Source Nodes
12
Control the Source Reporting Rates

• Coupling the application and the transport
  protocol
• Application layer determine each source nodes
  reporting rate with an optimization approach
• Justify/Ensure reliable data transport
• Minimize energy consumption to ensure reliable
  data transport
• Transport layer
• Provides the sink end-to-end communication cost
  from each source to the sink
• Minimize energy consumption in sensor-to-sink
  data communications
• Feed back reporting rates determined by the
  application

Sink
Source Nodes
13
Presentation Outlines

• 1. Introduction
• 2. Design Considerations
• 3. Protocol Implementation
• 4. Simulation Results
• 5. Conclusion

14
Protocol Requirements

• A good sensor-to-sink communication cost
  estimation mechanism
• A good routing scheme to achieve energy
  efficiency as well as in-network congestion
  avoidance.
• A feedback mechanism to adjust each sources
  reporting rate

15
Communication Cost Estimation

• Node Price (NP)
• A nodes node price is the energy consumed by all
  the in-network nodes for each packet successfully
  delivered from the node to the sink

Sink
16
Node Price Calculation

• Calculated in a backward propagating way
• The node prices of a nodes possible downstream
  neighbors
• Obtained by the feedbacks of its downstream
  neighbors
• The energy consumed to send packet to each
  downstream neighbor
• Calculated with link loss rate to each downstream
  neighbor
• The proportion of traffic the node sends to each
  downstream neighbor
• Determined by its routing scheme

See Description
17
Node Price Calculation
Sink
Downstream Neighbors
Weighted Average
Upstream Neighbors
Back
18
Node Price Calculation

• Link loss rate
• Mainly caused by three factors
• Congestion
• Signal Interference
• Fading.
• Packet loss rate will exhibit graceful increasing
  behavior as the communication load increases
  (IEEE 802.11 MAC)
• Reasonable to estimate the packet loss rate based
  on an EWMA (Exponential Weighted Moving Average)
  approach.

Current Estimation (1 - a) ? Current Loss a
? Previous Estimation
19
Node Price Calculation

• Estimate the link loss rate to each downstream
  neighbor
• Accurate and current link loss rate estimation
• Well indicates the congestion condition
• Well indicates the weak link
• Node Price based on loss rate estimation
• well indicates the dynamic wireless communication
  condition from the node to the sink
• can help to determine the reporting rates
• can help to determine the routing scheme

20
Node Price Calculation

• Link loss rate estimation
• Measured according to packet serial numbers holes
• Estimated with an EWMA approach.

SN 109
SN 100
Measured Loss Rate 2 / (109 - 100 1) 20
21
Routing Schemes

• Minimizing local node price.
• A node should minimize the energy consumed for
  the network to successfully deliver a packet to
  the sink from the node

Sink
Downstream Neighbors
Select the proportion of traffic routed to each
neighbor so that my node price is minimized
22
Routing Schemes

• Oscillation Avoidance

Congestion
Congestion
23
Routing Schemes

• Oscillation Avoidance
• Gradually shift traffic to best path
• Adaptive to downstream dynamics

Traffic proportion shifted to a better node
(with lower NP) at each time is
24
Diagram of PORT
25
Presentation Outlines

• 1. Introduction
• 2. Motivations and Design Considerations
• 3. Protocol Implementation
• 4. Simulation Results
• 5. Conclusion

26
Simulation Settings

• Coding PORT over NS-2
• Simulation Settings

27
Simulation Networks
28
Simulation Results

• Results

Total Energy Consumption (J)
29
Simulation Results

• Results

Total Energy Consumption (J)
30
Presentation Outlines

• 1. Introduction
• 2. Motivations and Design Considerations
• 3. Protocol Implementation
• 4. Simulation Results
• 5. Conclusion

31
Conclusion

• We discuss what a reliable transport protocol for
  wireless sensor network should do.
• We propose PORT, a price-oriented reliable
  transport protocol to address the reliable
  sensor-to-sink data transport issue
• PORT optimizes the energy consumptions with two
  schemes.
• The sink's optimization scheme that feeds back
  the optimal reporting rate of each source.
• A routing scheme for in-network nodes according
  to the feedback of downstream communication
  conditions to achieve energy-efficiency and avoid
  congestion.
• Simulation results in an application case study
  verify its effectiveness in reducing energy
  consumption.